I searched the entire twenty pages of Synchron topics to make sure this hasn't been covered, so I apologize if I somehow missed it, but the Synchron Celesta seems way too quiet no matter what I do with various settings, and this is coming from someone who typically aims for -16 dB for peaks and -20 to -24 dB for averages, to avoid too much trimming during mixing (most people I know record close to 0 or -6 dB; old school from 16-bit days).
Although I still need to use the original VSL percussion now and then for certain playing styles and contexts, I have fairly successful transferred much of my work over to Synchron Percussion, after carefully setting up my own presets. The Celesta is the one that eludes me so far, even though I feel it should slightly better than the original one. Although I could add gain to it later in the chain, I feel something must be wrong that it is so low, so am holding off on using it at all.
Stranger still, I have one part that has a few places where there are rapid staccato leaps, and other places where diads or triads are played in a semi-sustained dotted rhythm, and the "normal" playing style is where the volume drops precipitously even if I have mics maxed out, CC7 and CC11 at 127, and MIDI velocity near peak, or even with sustain pedal on, and longer note lengths. But the staccato parts come out loud and clear!
Using the original library, I had a keyswitch track to go between staccato and sustain/normal. That is no longer offered in the Synchron version, but I figure it wasn't necessarily essential in the original either. All the other pewrcussion elements seem equivalent in articulations and other aspects, between the original and the new Synchron library, once one studies more deeply the new system in Synchron Percussion. Celesta seems an exception.
If a different choice was made for some reason, that's fine, but I feel I MUST be missing something -- especially as no one else has reported this. It's true that some have reported the overall low output, but as I said, this one is significantly lower than ALL of the others, even when tweakable parameters are set to effectively maximize output.
Another interesting factor in my specific part, is that the staccato leaps are in the highest octave, and I do notice that this new version seems to get softer the lower it goes, but I thought the same celesta model was sampled this time around. If not, that might explain it, because I have two dozen cleesta libraries, most based on different models, and am aware that this instrument varies significantly across the model range.